Share your thoughts, 1 month free Claude Pro on usSee more

Visual Question Answering on GQA (test)

89.3Accuracy

Humans

Updated 3d ago

Evaluation Results

Method
Humans 2020.04	89.3	91.2	87.4	-	-	-	-
Kakao* 2019.07	73.33	-	-	-	-	-	-
270 2019.07	70.23	-	-	-	-	-	-
NSM 2019.07	67.25	-	-	-	-	-	-
LLaVA-13B 2024.05	67.1	-	-	-	-	-	-
Dense Connector 2024.05	66.6	-	-	-	-	-	-
LLaVA-NeXT 2024.05	65.4	-	-	-	-	-	-
Dense Connector 2024.05	65.4	-	-	-	-	-	-
Upper Bound 2026.04	65.4	-	-	-	-	-	-
Dense Connector 2024.05	65.1	-	-	-	-	-	-
LLaMA-VID 2024.05	65	-	-	-	-	-	-
MATA 2026.01	64.9	-	-	-	-	-	-
ShareGPT4V 2024.05	64.8	-	-	-	-	-	-
MATA 2026.01	64.7	-	-	-	-	-	-
MobileVLM V2 2024.05	64.6	-	-	-	-	-	-
Dense Connector 2024.05	64.6	-	-	-	-	-	-
Dense Connector 2024.05	64.4	-	-	-	-	-	-
Dense Connector 2024.05	64.3	-	-	-	-	-	-
LLaVA-NeXT-7B 2025.09	64.2	-	-	-	-	-	-
CLASP 2026.04	64.2	-	-	-	-	-	-
Finetuned mPLUG-Owl 2023.06	64	-	-	-	-	-	-
Dense Connector 2024.05	64	-	-	-	-	-	-
Dense Connector 2024.05	63.9	-	-	-	-	-	-
Dense Connector 2024.05	63.9	-	-	-	-	-	-
InternVL3.5 2026.01	63.8	-	-	-	-	-	-
LLaVA-LLaMA3 2024.05	63.5	-	-	-	-	-	-
Mini-Gemini 2024.05	63.4	-	-	-	-	-	-
LOVA³-7B 2024.05	63.4	-	-	-	-	-	-
LLaVA-v1.5 2024.05	63.3	-	-	-	-	-	-
VILA 2024.05	63.3	-	-	-	-	-	-
NSM 2020.04	63.2	78.9	49.3	-	-	-	-
CuMo 2024.05	63.2	-	-	-	-	-	-
CLASP 2026.04	63	-	-	-	-	-	-
LXRT 2019.07	62.71	-	-	-	-	-	-
SparseVLM 2026.04	62.7	-	-	-	-	-	-
ChatSearcher 2024.10	62.5	-	-	-	-	-	-
Qwen2.5-VL 2026.01	62.4	-	-	-	-	-	-
InternVL3 2026.01	62.4	-	-	-	-	-	-
IVM-Enhanced LLaVA-7B 2024.05	62.2	-	-	-	-	-	-
ZOO-Prune 2025.09	62.19	-	-	-	-	-	-
Qwen2-VL-Instruct 2025.05	62.18	-	-	-	-	-	-
Qwen2-VL-7B-GRPO-8k 2025.05	62.04	-	-	-	-	-	-
InstructBLIP 2023.06	62	-	-	-	-	-	-
LLaVA-7B 2024.05	62	-	-	-	-	-	-
LLaVA-1.5 2024.05	62	-	-	-	-	-	-
LLaVA-1.5-7B 2024.10	62	-	-	-	-	-	-
LLaVA-1.5-7B 2025.03	61.9	-	-	-	100	-	-
CLASP 2026.04	61.8	-	-	-	-	-	-
DivPrune 2025.09	61.58	-	-	-	-	-	-
Dense Connector 2024.05	61.5	-	-	-	-	-	-
InternVL2.5 2026.01	61.5	-	-	-	-	-	-
olmOCR-7B-0225-preview 2025.05	61.48	-	-	-	-	-	-
TSV Merging 2025.05	61.4	-	-	-	-	-	-
Iso-C 2025.05	61.34	-	-	-	-	-	-
Mixture Training 2025.05	61.33	-	-	-	-	-	-
TinyLLaVA 2024.05	61.3	-	-	-	-	-	-
VisionZip 2025.09	61.3	-	-	-	-	-	-
OptMerge 2025.05	61.29	-	-	-	-	-	-
GRN 2019.07	61.22	-	-	-	-	-	-
IPCV+FastV 2025.12	61.2	-	-	-	-	89.9	0.644
TwigVLM 2025.03	61.2	-	-	-	99.2	-	-
TwigVLM++ 2025.03	61.2	-	-	-	99.6	-	-
TA w/ DARE 2025.05	61.15	-	-	-	-	-	-
MobileVLM V2 2024.05	61.1	-	-	-	-	-	-
MSM 2019.07	61.09	-	-	-	-	-	-
ZOO-Prune 2025.09	60.97	-	-	-	-	-	-
Qwen2-VL-7B-Pokemon 2025.05	60.96	-	-	-	-	-	-
Task Arithmetic 2025.05	60.95	-	-	-	-	-	-
DREAM 2019.07	60.93	-	-	-	-	-	-
Individual VQA 2025.05	60.91	-	-	-	-	-	-
SparseVLM 2026.04	60.9	-	-	-	-	-	-
SK T-Brain* 2019.07	60.87	-	-	-	-	-	-
Vanilla 2025.12	60.8	-	-	-	-	100	1
DART (ViT) 2025.12	60.8	-	-	-	-	88.1	0.639
TwigVLM++ 2025.03	60.8	-	-	-	99.2	-	-
PKU 2019.07	60.79	-	-	-	-	-	-
TIES w/ DARE 2025.05	60.78	-	-	-	-	-	-
TwigVLM 2025.03	60.6	-	-	-	98.7	-	-
TIES Merging 2025.05	60.54	-	-	-	-	-	-
SparseVLM 2025.09	60.3	-	-	-	-	-	-
ToMe 2025.12	60.2	-	-	-	-	82.7	0.617
WUDI Merging 2025.05	60.11	-	-	-	-	-	-
VisionZip‡ 2025.03	60.1	-	-	-	98.3	-	-
Finetuned mPLUG-Owl (Pseudo Instruction) 2023.06	60	-	-	-	-	-	-
Task Arithmetic 2025.03	59.98	-	-	-	-	-	-
AdaMMS 2025.03	59.96	-	-	-	-	-	-
Musan 2019.07	59.93	-	-	-	-	-	-
MetaGPT 2025.03	59.93	-	-	-	-	-	-
ZOO-Prune 2025.09	59.93	-	-	-	-	-	-
ToFu 2025.12	59.8	-	-	-	-	83.4	0.619
TwigVLM++ 2025.03	59.7	-	-	-	97.7	-	-
DivPrune 2025.09	59.63	-	-	-	-	-	-
CogVLM(base) 2025.03	59.43	-	-	-	-	-	-
SparseVLM 2026.04	59.4	-	-	-	-	-	-
VisionZip 2025.09	59.3	-	-	-	-	-	-
Qwen-VL 2024.10	59.3	-	-	-	-	-	-
VisionZip 2025.03	59.3	-	-	-	97.4	-	-
VisionZip‡ 2025.03	58.9	-	-	-	97.3	-	-
DARE-Linear 2025.03	58.8	-	-	-	-	-	-
TwigVLM 2025.03	58.8	-	-	-	96	-	-

Showing 100 of 226 rows